The volatility risk premium
Selling options means many small wins and rare, enormous losses: the same profile as a grid. The difference is that here something real changes hands — a premium somebody pays to transfer a risk. This module is about telling when that profile is an insurance business and when it is just a deferred bill.
By the TradingCalculator.Pro team · Updated on · About us
Start your 7-day free trial →What you will learn
What it is
Implied volatility — the one baked into option prices — tends to trade above the volatility the asset ends up realising. That systematic gap is the premium: the option seller collects it in exchange for taking on the risk that realised volatility, this time, comes in higher.
Why it gets paid
It is not an inefficiency somebody forgot to arbitrage away: it is the price of insurance. A fund that needs not to blow up pays for protection even when it is expensive on average, the same way people buy car insurance knowing the insurer profits. The premium exists because there is structural demand for protection, which is why publishing it does not make it vanish.
What the evidence says, and with what caveats
The CBOE's systematic option-writing indices on the S&P 500 have historically shown a return similar to the index with considerably less volatility, which improves the risk-adjusted ratio. Two caveats that are not minor: those indices are computed GROSS, before transaction costs, and Sharpe barely penalises skew, so it flatters precisely the left-tail strategies. Treat it as a well-documented direction, not as a number you can carry to your own account.
The bill: 5 February 2018
The VIX had its largest one-day percentage jump on record that day, and the inverse volatility ETN XIV lost around 96 % of its value in a single session; its issuer announced liquidation the next day. Years of steady returns wiped out in hours. The lesson is not that the premium was not real: it is that those steady returns WERE the payment for carrying tail risk, and the bill arrives whole and at once.
How it differs from a grid
In profile, not at all: many wins, rare and large losses. In substance, entirely. Selling premium you collect something somebody willingly pays to transfer you a risk; in a grid you collect nothing from anybody, you merely defer recognising a loss. That is why one can have positive expectancy with management and the other has none with any.
How to run it without blowing up
Three non-negotiable rules. The stop: without it, the potential loss on a short-premium position is effectively unlimited, and that flips the expectancy. The size: a very small fraction, sized to survive the tail loss, not the average one. And correlation: ten short-volatility positions are not ten bets, they are one, and they sink together — the limit has to be on aggregate exposure, not per position.
What to look at in your own numbers
Win rate and Sharpe will not warn you: both look fine right up to the hit. The ones that do warn are the skewness, kurtosis and tail ratio in the journal. Strongly negative skew with a tail ratio below 1 means your extreme losses already exceed your extreme wins, and that is information about what is going to happen, not about what happened.